Android/frameworks_base
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

am 50cf95fc: am 87c6f9a5: Merge "This is wrong since it fixed the symptom not the problem. Alan is uploading a correct patch." into jb-mr1-dev

Browse Source 
* commit '50cf95fc7d5481d7f1d9a2c6bb7496a77ad9e926':
  This is wrong since it fixed the symptom not the problem. Alan is uploading a correct patch.
This commit is contained in:
Svetoslav Ganov
2012-10-11 17:23:43 -07:00
committed by Android Git Automerger
parent fdffe981fb 50cf95fc7d
commit 78c12209b6
1 changed files with 1 additions and 12 deletions
Download Patch File Download Diff File Expand all files Collapse all files

13
core/java/android/webkit/AccessibilityInjector.java
View File

@@ -710,18 +710,7 @@ class AccessibilityInjector {
"javascript:(function() { %s.onResult(%d, %s); })();";
// Time in milliseconds to wait for a result before failing.
// Based on recorded times, we have found that in a complex real-world
// web app (such as the desktop version of Gmail), there can be spikes
// of ~2600ms in the worst case. These are temporary spikes and are not
// repeatable; GMail eventually settles down to around ~60ms. The
// longest duration that is consistently repeatable is ~300ms when
// loading extremely large plain text documents in WebView.
// If this timeout hits, the navigation is considered to have "failed",
// meaning there is no content. Since the longer spikes are one-off
// events triggered by the page loading and itself running a large
// amount of JS, subsequent runs would succeed, so the worst impact
// is that the first run will look like it had not loaded yet.
private static final long RESULT_TIMEOUT = 1500;
private static final long RESULT_TIMEOUT = 5000;
private final AtomicInteger mResultIdCounter = new AtomicInteger();
private final Object mResultLock = new Object();